On May 11, 2021, an inspector checking the Hernando de Soto Bridge from above the deck spotted something no routine inspection expects to find. One of the two main steel tension ties in the eastbound arch span had fractured across nearly half its cross-section. The girder measured 26 by 33 inches, and the fracture had severed the top flange and outer side plate before extending across nearly half of the bottom flange.
The bridge normally carries about 41,000 vehicles daily between Memphis and West Memphis and is one of only two major Mississippi River crossings in the area. What makes the story worth returning to is not just the fracture that was seen. It’s the twelve others that weren’t.
Why one split tie can take a whole span down
The Hernando de Soto is a tied-arch bridge, and that design puts enormous trust in its steel tension ties. In a tied arch, the arch wants to push its supports outward. The tie running along the bottom resists that thrust in tension. These members were classed as fracture critical, meaning failure could threaten the stability of the span.
The engineers who examined the Memphis fracture did not hide their surprise. Frank Russo put it this way: “Usually when something like this breaks, you find it completely broken. It’s very remarkable that this member didn’t completely fracture.” Aaron Stover was blunter when he first saw the photographs, asking simply, “How is this bridge still standing?” That was one engineer’s startled reaction, not a calculation that collapse was imminent, but it explains the urgency.
So where did the fracture come from? An investigation by Wiss, Janney, Elstner Associates concluded that it likely began in a fabrication weld repair made before the bridge opened. Hydrogen-assisted cracks were already present in the repaired weld, and the eventual fracture developed in several stages over years. The underlying defect was nearly as old as the bridge itself.
The precedent that reads like a rehearsal
This was not the first tied-arch bridge pulled from service over the same kind of problem. On September 9, 2011, the Sherman-Minton Bridge carrying I-64 over the Ohio River between Louisville and New Albany closed after inspectors found a critical crack in a tension tie. It remained closed for roughly five months.
The two bridges had an uncomfortable amount in common. Both were tied-arch structures, and both used high-strength T-1 steel fabricated before modern fracture-control requirements. In both cases, investigators traced the trouble to hydrogen-assisted cracking associated with welding during fabrication. FHWA later linked the cases when it ordered broader inspections.
What the eye kept missing
The unsettling part is that cracks like these can be hidden inside a weld. A surface can look normal even while defects remain beneath it, which is why ultrasonic and other non-destructive testing matters.
The Hernando de Soto had actually undergone ultrasound testing before. In 1982, Tennessee tested its welds, including the one that later fractured. Inspectors recorded a small defect there and classified it as slag, but the test missed two much larger hydrogen cracks nearby in the repaired portion of the same weld.
That is why the sweep after the 2021 closure mattered. Once crews used non-destructive testing on the other welds, the picture changed. FHWA recorded that testing found 12 other butt welds containing hydrogen cracking that visual inspection had not detected.
The rule that changed nationwide
After the Sherman-Minton closure, FHWA had already issued Technical Advisory 5140.32, recommending that bridge owners identify vulnerable T-1 steel members and, where cracking had already been found, verify butt welds with visual and non-destructive testing. Memphis pushed the agency further.
On December 13, 2021, FHWA issued an action memo requiring state transportation departments to identify older fracture-critical T-1 steel members and supplement hands-on inspections with non-destructive testing of their butt welds unless adequate previous testing could be documented. The memo warned that failure to act could put structures at risk of failure.
We’re not engineers, and we’ll leave the metallurgy to the people who ran the tests. What we take from the sequence is a plainer thing. A flaw put into a weld before the bridge opened in 1973 survived a 1982 ultrasonic inspection and carried traffic for decades before becoming a huge visible fracture. Testing after the closure then found twelve more affected welds that visual inspection could not see. The near-miss over the Mississippi is what finally turned that kind of machine-assisted inspection from federal advice into a requirement.